I lecture in a range of non-orthodox subjects at the University of Western Sydney, in Sydney, Australia. This page will provide a permanent link to these lectures, including PowerPoint slides, and screen recordings of the lectures.
Behavioural Finance
Lecture 01: Debunking Consumer Theory
The concept I teach here–Revealed Preference–was taught in 1st year 40 years ago, when I was an fresher undergraduate. But the tuition of Neoclassical economics has been so dumbed down over the years that most undergraduates don’t heard of it. I expect it’s reserved as punishment for those who undertake an Honors degree these days!
After outlining the theory, I then cover the excellent experimental disproof of the theory by the German economists Reinhard Sippel:
Sippel, R. (1997). “An Experiment on the Pure Theory of Consumer’s Behaviour.” The Economic Journal 107 (444): 1431-1444.
I’m sure that disproving the theory wasn’t Sippel’s original intention. Instead, I suspect that he undertook the experiment to show to his students that their “indifference curves” could be inferred from their purchases, as Samuelson claimed long ago when he dreamed up the concept of Revealed Preference:
Samuelson, P. A. (1938). “A Note on the Pure Theory of Consumer’s Behaviour.” Economica 5 (17): 61-71.
Instead, Sippel found that his experimental subjects violated the “Axioms of Revealed Preference”.
In the first half of the lecture, I cover the axioms of revealed preference, and Sippel’s results:
In the second half, I interpret these results using ideas from computation theory:
Lecture 02: Debunking Supply Theory
In the first half of this lecture, I show that even if all consumers were utility maximizers whose individual demand curves obeyed the “Law of Demand”, the market demand curve derived from aggregating these consumers could have any shape at all. This result, known as the “Sonnenschein-Mantel-Debreu Conditions“, is actually a Proof by Contradiction that market demand curves do not obey the “Law” of Demand, and therefore that Marshallian partial equilibrium modeling of individual markets is invalid–let alone the Neoclassical practice of modeling the entire macroeconomy as a single agent in “Dynamic Stochastic General Equilibrium” models.
In the second half, I show that even if the market demand curve were valid, supply and demand analysis is still impossible. A supply curve that is independent of the demand curve can only be derived if firms set price equal to marginal cost, which neoclassical economists claim is the consequence of profit-maximizing behavior by competitive firms. I show: that equating marginal cost and marginal revenue does not maximize profits; that therefore if firms from different industry structures faced the same costs, the amount produced is independent of the number of firms in the industry and corresponds to the so-called “monopoly” level of output. In this case, a supply curve cannot be derived.
Here are the Powerpoint files for this lecture: Part 1; Part 2.
Lecture 03: Debunking Finance Theory
CAPM still dominates the teaching of finance, but it was always nonsense because it relied on the following two assumptions:
“In order to derive conditions for equilibrium in the capital market we invoke two assumptions.
“First, we assume a common pure rate of interest, with all investors able to borrow or lend funds on equal terms.
Second, we assume homogeneity of investor expectations:
investors are assumed to agree on the prospects of various investments—the expected values, standard deviations and correlation coefficients described in Part II.”
Even Sharpe had to instantly admit that, as assumptions go, these two were doozies: he continued that ”Needless to say, these are highly restrictive and undoubtedly unrealistic assumptions.” But never let realism get in the way of a neoclassical theory! He defended this nonsense with a twisted appeal to Milton Friedman’s “instrumentalist” methodology:
“However, since the proper test of a theory is not the realism of its assumptions but the acceptability of its implications,
and since these assumptions imply equilibrium conditions which form a major part of classical financial doctrine,
it is far from clear that this formulation should be rejected—especially in view of the dearth of alternative models leading to similar results.” (Sharpe 1964, pp. 433-434)
A major facet of CAPM was using the concept of Expected Value developed by John von Neumann and Oskar Morgenstern in Theory of Games and Economic Behavior. As is so often the case in economics, their work was misinterpreted. While CAPM involved blending neoclassical indifference curve analysis of individual behavior–which I debunked in the first lecture in this series–with von Neumann’s Expected Value analysis, von Neumann’s intention was to develop a numerical measure of utility and eliminate indifference curves from economic theory.
Despite these obvious reasons to dismiss CAPM, it took over the profession, and part of the reason for its success was its apparent close fit to the data at the time it was developed. But was this just a fluke, given the tranquil economic times during which CAPM was developed?
Even Behavioral Finance, which is critical of CAPM, has largely been developed on a misunderstanding of von Neumann. It applies the subjective concept of utility and expected returns to argue that there are numerous “paradoxes” where people don’t behave “rationally” given a financial choice, and that these “paradoxes” can’t be explained by risk aversion, loss aversion, or all other manner of conventional explanations.
In fact all these paradoxes disappear if objective probability is used, which is what von Neumann and Morgenstern insisted upon–literally–in their book:
“Probability has often been visualized as a subjective concept more or less in the nature of estimation. Since we propose to use it in constructing an individual, numerical estimation of utility, the above view of probability would not serve our purpose. The simplest procedure is, therefore, to insist upon the alternative, perfectly well founded interpretation of probability as frequency in long runs.” (von Neumann & Morgenstern 1944, p. 19)
Finally, economists normally deride the “payback period” means of deciding between investments, on the argument that this ignores the time value of money–an argument you’ll find in the Wikipedia entry on this topic. But John Blatt showed in in the 1980s that the payback period takes account of both the time value of money and uncertainty about the future, at least to a rudimentary level.
Here are the two Powerpoint Files for this lecture (Part 1; Part 2).
[...] this stuff (http://debunkingeconomics.com/lectures/) from Keen that mathematically proves that the theoretical foundation of neoclassical economics is [...]
March 1, 2012 at 10:52 pm
I have been enjoying watching your videos and have just moved on to the lectures, which I’m enjoying. I’m a computer science professor rather than an economics student, but they’re quite enlightening nonetheless. As a computer science professor, after watching your video, I do feel a need to make a couple of comments. The first is that the sort you described (and that was illustrated on the slide) is not a Bubble Sort, it is a Quick Sort. Everything else you said about it was correct for Quick Sort, but most of what you said was not accurate for Bubble Sort.
I should also warn you about talking too strongly about the strength of passwords. It’s certainly correct that the number of possible passwords which can be chosen which are n characters long is exponential in n, so your basic point is correct, but when we examine how people actually choose passwords, we find that they’re very biased in their selections and not very good at choosing or remembering random strings of characters. This means that a large number of passwords can be guessed without having to exhaust more than a very small fraction of that exponential search-space. This is called a “dictionary attack” and is quite commonly used. Also, note that the common password length is only about 8 lowercase characters, so even with an exponential scale, if the exponent is small, then it can be guessed feasibly. 26^8 is 208 thousand million, but when you have computers which can do several thousand million calculations per second the time needed to find the right password becomes quite small. This is why longer and more varied passwords are recommended, but recommended doesn’t mean used. In theory, passwords can be quite strong, but, in practice, they’re often not.
March 1, 2012 at 11:02 pm
Ah, thanks Keith! I sometimes make stuff-ups like that when I use analogies or techniques that are not part of my everyday fare.
I learnt about the Quicksort routine (and Bubble sort) when self-teaching myself programming 30+ years ago. Obviously when I reached for an instance of a computer algorithm, I chose the wrong name. I’ll fix that next time I give that lecture.
Ditto the point about passwords. Sometimes however I have to just provide the generalities of an issue when using it as an example in a lecture; but footnotes exist to clarify such things in documents, and as I develop the lectures I hope I can add those qualifications via footnotes or hyperlinks. I’ve done quite a bit of that in the most recent edition of Debunking Economics.